LiveXP is a very flexible PE 1x, using Explorer as shell, supporting SATA/IDE drives, USB devices, and many other features and tweaks. This project is based on PE environment and will use a Windows XP/2003 Install CD.
Download server link to use with the WinBuilder Download Centre: www.livexp.galapo.net.
Download the most recent archive of the project: 
3 Pages
1 2 3 >
|
 |
LiveXP slow to boot from USB hard drive, Taking over 10 minutes to boot |
|
|
|
Feb 5 2010, 11:49 AM
|
|
|
|
I recently built my first LiveXP CD after downloading the Live XP (current) project from www.winbuilder.net Since then I have learned how to add/remove components with much success.
I have also been able to make my project run from a UFD through using PE2USB and instructions elsewhere on this forum.
As part of that process I successfully modified ntdetect.com to create the 47,596 byte version and copied setupldr.bin to NTLDR. The UFD boots OK but it is much slower than booting via the CD.
I hunted down an old 15gb hard drive and decided to see if I could make LiveXP run from this. The drive was partitioned to create a 1.9gb active primary FAT partition named LiveXP, with the remainder going to a NTFS logical. I copied I386 and Program Files folders, and ModelRam.exe from E:\Winbuilder\Target\LiveXP \ to the root of the USB HDD's 1.9gb FAT partition and renamed the I386 folder to minint I then copied across the modified ntdetect.com and NTLDR. Much to my surprise the drive booted at the first attempt but it was painfully slow, taking over 10 minutes before completion.
The CD, UFD and USB HDD are based upon an identical project, which creates a 180mb LiveXP.ISO The CD version takes around 2 minutes to boot. The UFD is very slow compared to this, but the USB HDD is the worst. I would have expected the opposite.
Am I possibly doing something wrong, or is very slow booting from external hard drives normal?
I would be very appreciative of any feedback upon this subject.
|
|
|
|
Feb 5 2010, 12:26 PM
|
|
|

|
|
Joined: 8-January 10
From: The Outside of the Asylum
Posts: 2,277
Thank(s): 252
 Italy

|
|
Most probably it mainly depends on your hardware. Basically: - some motherboards have ONLY USB 1.1 CHIPs
- some motherboards have USB 2.0 Chips, BUT have only USB 1.1 speed when booting from USB
- some motherboard have BOTH USB2.0 chips AND support for USB 2.0 speed when booting
http://en.wikipedia.org/wiki/Universal_Serial_BusQUOTE The full speed rate of 12 Mbit/s (~1.43 MB/s) is the basic USB data rate defined by USB 1.1. All USB hubs support full speed. A hi-speed (USB 2.0) rate of 480 Mbit/s (~57 MB/s) was introduced in 2001. All hi-speed devices are capable of falling back to full-speed operation if necessary; they are backward compatible. Connectors are identical. Check this also: http://en.wikipedia.org/wiki/USB_flash_driveReal world speeds are much lesser than that, a typical USB Flash stick will be between 8 and 20 MB/s, and average USB hard disk controller will be about 30 MB/s on a USB 2.0 bus. Some tests performed/related threads: http://www.boot-land.net/forums/index.php?...=9347&st=15http://www.911cd.net/forums//index.php?showtopic=21131http://www.911cd.net/forums//index.php?showtopic=21965Check your hardware, first thing. There are two possible workarounds for situation #2 (and NO workarounds for case #1): - using PLoP as boot loader
- using XP Kansas City Shuffle (often called "Fake signature" or "kicker-boot")
If you are on situation #3, then something else is the cause. Other considerations: http://www.msfn.org/board/index.php?showtopic=125116 Wonko
--------------------
|
|
|
|
Feb 5 2010, 01:11 PM
|
|
|
|
Wow - that's a pretty comprehensive response - thank you!
My P45 motherboard is fairly new and includes the Intel ICH10R Chipset, which apparently provides a Hi-Speed USB (USB2.0) controller, 480Mb/sec, up to 12 ports. I regularly use USB HDs for backup and imaging purposes and it usually seems fairly speedy (by USB standards anyway). It takes 10 seconds to write a 300mb file to it, and 6 seconds bring it back. The UFD equivalents are 110 seconds and 15, so the USB HD is definitely the quicker of the two options.
These figures suggest that booting LiveXP from the USB HD should be faster, whereas the opposite appeared to be the case. This makes me question whether I am missing some vital step, or whether the project build should differ for CDs, UFDs, and USB HDs.
It's not the end of the world, and I only really did it because 'it was there'. I was just very surprised that it should take over 10 minutes to boot, when the USB HD only holds 190mb of information, with a good chunk of that being dedicated to program files.
|
|
|
|
Feb 5 2010, 02:37 PM
|
|
|

|
|
Joined: 8-January 10
From: The Outside of the Asylum
Posts: 2,277
Thank(s): 252
 Italy

|
|
Additionally, always use USB ports in the back (the ones directly soldered to the mottherboard) and NOT the ones in the front, as often the latter ones are slower/badly working. And you did not state the filesystem you used.  If the ports boot at USB 2.0 speed it is possible that something else is slowing the booting of LiveXP abnormally. ("botched LiveXP build) To make a comparison, create an UBCD4WIN USB stick: http://www.ubcd4win.com/http://ubcd4win.com/forum/index.php?showtopic=11375If the UBCD4WIN "normal" boot takes approximately the same time, your LiveXP building is OK and your USB bus is not. (and you should get a MUCH faster boot using the "kicker") Wonko
--------------------
|
|
|
|
Feb 5 2010, 03:37 PM
|
|

|
|
Joined: 1-February 07
From: California
Posts: 1,744
Thank(s): 280
 United States

|
|
Hi Sool I've had the best luck using the great USB setup tools by wimb described here, the names can be a bit misleading (not just for installing OS etc.), I can get most USB drives to boot in under 2 min. even on older machines; UFD is best, usually coming in under 30 sec. There is a bit of information on it here under 'More Advanced Options' as well
--------------------
|
|
|
|
Feb 5 2010, 05:45 PM
|
|

|
|
Joined: 29-December 06
Posts: 5,256
Thank(s): 156

|
|
QUOTE (Sool @ Feb 5 2010, 02:11 PM)  It takes 10 seconds to write a 300mb file to it, and 6 seconds bring it back. The UFD equivalents are 110 seconds and 15, so the USB HD is definitely the quicker of the two options.
These figures suggest that booting LiveXP from the USB HD should be faster, whereas the opposite appeared to be the case. HDD are faster when reading sequential files (i.e. big files), reading lots of little files - USB-Sticks are faster! You can use a tool like HD-Tach, Atto disk or HDTune to check the performance of your stick in the booted LiveXP.
--------------------
|
|
|
|
Feb 5 2010, 06:02 PM
|
|
|
|
Thanks for the input guys!
I am using XP on a fairly high spec quad core pc with the NTFS file system and understand the difference between the front and rear USB ports.
I shall play around over the weekend with your various suggestions and see where I end up. Ultimately, I envisage my bootable USB HD seeing little, if any, use and I need to bear this in mind before spending too much time on it. Quite simply it occurred to me that such a drive could one day prove useful, say as a one-stop solution for recovering data onto the spare partition from a non-booting PC or somesuch. Obviously there are numerous alternative workarounds to such a situation, but I thought I'd have a go anyway.
I'm still finding my feet with Winbuilder, and suspect that I may have become confused by some of the references to BootSDI and WimBoot (neither of which I am using) that I have seen in the tutorials, with some posts suggesting that differing builds are needed, depending upon whether they are aimed at UFD, USB HD, and CDs.
Although I have played with adding and removing some of the applications, I have essentially left the downloaded 'Live XP (current) project' alone, when it comes to the core operations, which hopefully rules out the possibility of a botched build.
Ultimately, I found myself a little surprised that a build that boots via CD in 2 minutes, takes five times longer when booting from a USB device. I am now starting to question whether this is perhaps normal.
|
|
|
|
Feb 5 2010, 06:15 PM
|
|

|
|
Joined: 29-December 06
Posts: 5,256
Thank(s): 156

|
|
USB1.1 read speed is about 1,5MB/s CD read speed is 150kB/s x the factor mentioned on the drive. Do the math and you will find a 11x dive to be already faster. Because of greatly different access times, a direct comparison does not make much sense, but to give you an idea.
--------------------
|
|
|
|
Feb 5 2010, 06:16 PM
|
|
|

|
|
Joined: 8-January 10
From: The Outside of the Asylum
Posts: 2,277
Thank(s): 252
 Italy

|
|
QUOTE (Sool @ Feb 5 2010, 07:02 PM)  Ultimately, I found myself a little surprised that a build that boots via CD in 2 minutes, takes five times longer when booting from a USB device. I am now starting to question whether this is perhaps normal. Ok, let's clear this point, IT IS NOT "normal", if the device and bus operate at USB 2.0 speed. (unless you have an EXTREMELY fast CD drive  )  Wonko
--------------------
|
|
|
|
Feb 5 2010, 08:16 PM
|
|
|
|
Thanks again!
So it isn't normal. No wonder I'm scratching my head (or tearing my hair out to be nearer to the truth). So...
1. A direct copy of the CD's I386 (renamed minint) and Program Files folders, along with Modelram.exe, the modified ntdetect.com and NTLDR, should theoretically boot at a similar (or faster) speed from a USB device., irrespective of the various workarounds suggested.
2. It has nothing to do with utilising BootSDI or WimBoot (neither of which I'm using). Even though amalux's March 2008 tutorial suggests the UFD builds need to be based upon BootSDI
3. There are no extra boxes to tick, or files to include for USB builds.
4. It has nothing to do with adding Win 2k3sp1 files (mentioned by amalux but contradicted by? elsewhere).
5. I am really struggling to blame my otherwise speedy hardware.This suggests that there must be something significantly wrong with my build, even though this is largely the same as the downloaded project.
6. I am possibly best returning to that original project, or a really stripped one, and seeing how that behaves on USB
Unfortunately, all sorts of other things are going to get in my way over the next 36 hours. Knowing that what I am seeing is not normal, is a great help though.
Cheers everyone!
|
|
|
|
3 Pages
1 2 3 >
|
 |
1 User(s) are reading this topic ()
|
|